Research On Offloading Method Of Computing Task For Connected Vehicle Orient To 5G Internet Of Vehicle

With the arrival of the 5G era,"5G+ Internet of vehicles" is considered to be an indispensable technical support for the realization of autonomous driving or even driverless driving.The Internet of vehicles has been developing towards Internet connection and intelligence,from single-vehicle intelligence to multi-vehicle collaboration,and the collaborative development of "smart cars" and "smart roads" will help ease traffic congestion,reduce the probability of accidents,and make it easier for people to travel.It can be predicted that car-road cooperation has a huge space for development in China.However,due to its limited computing and storage resources,intelligent connected vehicles cannot process a large number of computation-intensive and time-delay sensitive computing tasks in real time.Relying on the Internet of vehicles technology,connected vehicles can migration some computing tasks to edge servers in the Internet of vehicles for processing.In this paper,an improved chaotic differential evolution algorithm is proposed based on the detailed analysis and research of the computing task migration method of the connected vehicle,and combined with the multi-attribute decision method,the optimization of the computing task migration method of the connected vehicle is realized.The correctness and effectiveness of the proposed migration method are verified by simulation.The specific research work of this research mainly includes the following contents:(1)In order to solve the problem of computing task migration in connected vehicle scenarios,this paper proposes a computing task migration method based on improved chaotic differential evolution algorithm.Based on the detailed analysis and research on the computing task migration of connected vehicles in typical scenarios of Internet of vehicles,a computing task migration model with multi-objective optimization of time delay and energy consumption was established,and the variable of computing task migration of Internet of vehicles was discussed and proved to be a discrete decision problem.The chaotic differential evolution algorithm has a good global search ability,but it is usually used to solve continuous decision problems.Therefore,the first improvement of the algorithm is carried out in this paper,and the continuous variables are discretized to solve the computational migration variables.(2)After the first improvement of the chaotic differential evolution algorithm,due to the high computational complexity and slow solving process of the non-dominated sorting method,it does not meet the strict requirements of the Internet of vehicles for time delay.Therefore,the second improvement of the chaotic differential evolution algorithm is made,in which the nondominated sorting method is improved to the fast non-dominated sorting method,which can reduce the computational complexity of the algorithm,improve the computational speed of the algorithm,and shorten the algorithm time.The simulation results show that the improved algorithm can meet the requirements of low time delay and low energy consumption when applied to networked vehicle computing task migration.(3)Aiming at the migration problem of multiple computing tasks of connected vehicles in the scenario of Internet of vehicles,the time-delay sensitive tasks are analyzed and studied based on the intensive computing tasks.Sort multiple time-sensitive computing tasks according to task priority.The improved chaotic differential evolution algorithm is used to solve the problem.The simulation results show this method meets the requirements of the vehicle network for delay.(4)In order to solve problem that the optimal solution is a group when improved chaotic differential evolution algorithm is used to solve optimal solution of task migration model,it is necessary to improve the improved chaotic differential evolution algorithm.Because multiattribute decision making algorithm can choose the best one in a group of optimal solutions,improved chaotic differential evolution algorithm is combined with the multi-attribute decision making algorithm to obtain improved hybrid evolutionary algorithm.The simulation show that compared with the improved chaotic differential evolution algorithm,the improved hybrid evolutionary algorithm has lower time delay and energy consumption.